LC problem, unstable

Tx loop 5W.
Series resonance circuit, 11.8 kHz, loop diameter 30cm.
C = 6267 pF, L = 29mH
Δ = parameter changes.
after 20 min voltage changes 3.6 - 2.6V , Δ U = 1V
Δ t°C = 30°C
Δ L =0.5mH
Δ C = 50 pF
Measurement taken after 20 min of transmission, power off, and measurement is made quickly (in 5 sec)
Δ L = 0.1 mH = 20 Hz frequency shift.
Δ R = 0.4 ohm
Δ C = 10pF = 9 Hz frequency shift.
Total shift 29Hz
To bring circuit to resonance again I have to add a capacitor 150pF = 140 Hz frequency shift.
140 - 29 = 111 Hz missing

I think you have an antenna and a question, but that is not clear for me. Is there a Arduino somewhere ? Perhaps you have more luck at StackExchange, but say in words what you have and what the problem is.

Arduino is used for generating 11.8kHz signal. The problem is instability of resonance circuit

Could you show a schematic, a photo and the sketch ?

If you put 5W into a coil, that changes the coil. You could put a coil in epoxy resin to compare the results.

I'm not convinced that 111 Hz is missing.
The measurements result into a change of 1% (11.8 kHz versus 111 Hz). How accurate can you measure the components ? You should measure them with 0.1%, then perhaps the change turns out to be less or be within the accuracy of the measurements.

Since you have an Arduino, you can use the Arduino to find the peak of the resonator frequency.

I have smaller coil with epoxy, 150mA*14V = 2.1W.
Frequency on start 9.28kHz, 2 hours later 9.07kHz, delta f =201 Hz = 0.25%
Schematic as below with changes:
1Ct connected to Arduino instead to ground, LC connected in series.

Yesterday evening capacitor has warm, today morning is cold, but frequency still changing.
0.25% - it is look like is in range of component changes, so what will be the way to stabilize LC resonance frequency
FullSizeRender 2
FullSizeRender

Just because your coil inductance changes this will not change the frequency of this circuit. Yes it will move the resonance frequency but that does not change the frequency it produces.

That is governed by the chip driving the resonant circuit. So it is you chip that is unstable. I have not met this chip before but it seems the frequency is being controlled by a simple R C circuit, and that is giving you temperature variations.

I would look into driving this chip with an external oscillator and making a crystal oscillator to drive it with.

Just looked at the data sheet and confirmed what I thought.

The other fix is to use a capacitor type that has a lower temperature coefficient. In radio circuits you can get a stable capacitor by combining types with positive and negative temperature coefficients so that they balance out and give a stable result over a wide temperature range.

Frequency is sable R C circuit is synchronized by Arduino, for that I connect Ct to arduino not to ground.

Then why was this not shown on the circuit?

So the problem is an unstable clock on the Arduino. Still the same solution, make it stable.

1 Like

That will change LC resonance frequency, to compensate that I have to add a 121pF capacitor

It says here

No you don’t seem to get it, driving a resonant circuit at a slightly different frequency will not change the frequency output. It will affect the current through the coil and the field it produces which will be slightly down, but the frequency out will always be the frequency you are driving the coil with.

I tried few different generators the results are the same.

Again, the frequency is stable but amplitude in RX antenna is changing by 1V, to make it the same as on beginning when power is turn on a capacitor has to be added = tuning LC circuit

Yes makes sense, more sense than your title.

The answer is, like I said, to make your resonant capacitor more temperature stable. What type of capacitor are you using? Have you looked up capacitor types and seen the temperature coefficients?

I am using CBB capacitor, I thinking about NPO capacitor, so far i didn't find 3.3nF/ 2000V, yes there is a high voltage on LC.
OIP.PI3OLHP4DjqSaXAJpNNpugHaHa

Yes I know that, glad you know it too.
Things you can try.
Find out the ripple current rating if the capacitor you are using and try and get one with a higher rating. Or use two or more smaller capacitors in parallel to spread the heating over several components.

Try and reduce the losses in the coil by using thicker wire and changing the core you wind it on.

I have 10 pcs CBB capacitors from one supplier, one of them is acceptable stable, the changes are 0.4V in less han 1 min then voltage is stop changing.
1 of 10 capacitor is ok

There is a nice smoke and burning smell when touching it by finger, the spark is there also.

Thanks for suggestions

Now I am using 4 capacitors instead of 1, bingo, Tx voltage is doubled, amplitude is stable in 2 min.

What is the project for ?
You could use the Arduino to find the right frequency during runtime, or the Arduino should be able to tune the circuit to fix the frequency.

You have now something that will drop the power with the slightest change.